At a glance
Nothing pushes a free particle, so it looks like the simplest case in mechanics. This sim builds a Gaussian wave packet from three numbers (k₀, σ and m) and follows it in time, as the wavefunction in the upper plot and as a probability density in the lower.
What you'll learn
- How a packet is built from plane waves, and what σ measures (it is not a width in x).
- How to read ψ and the probability density P(x,t) off the two plots.
- How the packet's centre and width depend on k₀, σ, m and the time t.
- How large the effect is for an electron, an atom and a grain of dust.
What's on screen
- Upper plot, blue: Re[√Cψ], the real part of the wavefunction.
- Upper plot, purple: Im[√Cψ], the imaginary part. √C is a constant, so P = (Re)² + (Im)².
- Lower plot, green: the probability density P(x,t) ≥ 0. Its total area is 1.
- Live Metrics: Time, xpeak (the centre of P), σ²|α|² (the width readout), Peak P, v = ħk₀/m and Norm (the area under P).
The controls
Press ⌖ on a row to see where that control sits in the sim.
| Control | What it changes | Range | |
|---|---|---|---|
| k₀ | The mean wavenumber of the packet, so its mean momentum ħk₀. The sign sets the direction. | −2.00 to 2.00 opens at 0.00 | |
| σ | The spread of momenta in the packet (the width of A(k), in k-space). | 0.70 to 2.00 opens at 1.00 | |
| m (mₑ) | The particle's mass, in electron masses. | 0.5 to 5.0 opens at 1.0 | |
| Time Speed | How fast the clock runs. The physics is the same at any speed. | 0.0 to 4.0 opens at 0.2 | |
| Re / Im (Display) | Shows or hides the blue and purple curves. P is unaffected. | on / off both on | |
| Live Metrics | Six readouts: Time, xpeak, σ²|α|², Peak P, v and Norm. | panel | |
| Play / Pause | Runs or freezes the clock. The sim opens paused at t = 0. | top bar | |
| Reset | Rewinds the clock to t = 0 and plays. | top bar |
Changing k₀, σ or m rebuilds the packet at t = 0. Fixed: ħ = 1, time is in the sim's own "a.u.", and the plots show x from −40 to 40. Each plot's vertical scale is set from the t = 0 state of the current k₀, σ, m, so a shrinking curve means a shrinking value.
A packet is a sum of plane waves
A plane wave eikx has momentum ħk and no position. A localised packet adds many of them, with Gaussian weights centred on k₀: A(k) ∝ exp[−(k − k₀)²/2σ²].
- σ is the spread in k, not in x. |A|² has standard deviation σ/√2, so Δp = σħ/√2.
- At t = 0 the sum is a smooth hump times a carrier wave: ψ(x,0) ∝ e−σ²x²/2 · eik₀x. The hump sets the shape of P. k₀ sets the ripples inside it.
- The hump has width Δx(0) = 1/(σ√2), so a wider A(k) means a narrower hump. Together, Δx(0)·Δp = ħ/2.
Every momentum runs its own clock
A free plane wave evolves as ei(kx − ħk²t/2m): its phase turns at the rate ħk²/2m, which depends on k. The packet is the sum, taken with the same weights A(k).
- xpeak is the centre of the hump. σ²|α|² is the width readout, equal to 2Δx².
- The prefactor 1/(σ√π|α|) falls as |α| rises, so the hump flattens while its area stays 1. Norm stays 1.00 unless the packet outgrows the ±40 window.
The centre moves at ħk₀/m
- Drift. The centre moves at v = ħk₀/m, the group velocity. That is the speed of a classical particle with momentum p = ħk₀. A negative k₀ moves it left.
- k₀ is absent from the width law. A moving packet spreads exactly as a resting one does.
- Check: with k₀ = 1 and m = 1, v = 1, so xpeak equals the Time readout. With m = 2, xpeak is half of that.
The width grows on the time scale τ
These sliders move the sim too.
- τ grows with m and falls with σ². A heavier packet spreads more slowly. A packet that starts narrower in x (larger σ) spreads faster.
- Since τ = 2m·Δx(0)²/ħ, halving the starting width cuts τ by four.
Momentum stays fixed, phases do not
- Each component keeps its weight |A(k)|, so the momentum spread Δp = σħ/√2 never changes. Only the phases turn.
- Components with larger k move faster (ħk/m), so they separate. At late times, Δx ≈ Δp·t/m.
- That sorting shows in ψ. For k₀ = 0 the local wavenumber is k(x) = ħtx/(m|α|²): small at the centre, larger toward the edges. At late times k ≈ mx/ħt.
- Δx·Δp ≥ ħ/2 is a floor. The packet sits on it only at t = 0.
Where it matters
- Electrons. Localised to 0.1 nm, an electron has τ = 2mΔx(0)²/ħ ≈ 1.7 × 10⁻¹⁶ s. A free packet that small would double σ²|α|² in about 0.17 fs.
- Cold atoms. A ⁸⁷Rb atom localised to 1 µm has τ ≈ 2.7 ms, a laboratory time scale.
- Everyday objects. A 1 g bead localised to 1 µm has τ ≈ 6 × 10¹¹ years, about forty times the age of the universe. Spreading is real, and unobservable.
Go deeper: the sim in atomic units
- The sim sets ħ = 1 and gives m in electron masses. It does not name a length unit, only "a.u." for time.
- If x is in bohr radii and time in atomic units (1 a.u. = 24.19 as), then σ = 1 and m = 1 give Δx(0) = 0.707 bohr ≈ 0.037 nm and τ = 1 a.u. ≈ 24 as.
Summary
Key points
- The packet is a sum of plane waves with Gaussian weights; σ is the spread in k, and Δx(0)·Δp = ħ/2.
- xpeak = ħk₀t/m. The width law σ²|α|² = 1/σ² + σ²ħ²t²/m² does not contain k₀.
- Δx(t) = Δx(0)√(1 + (t/τ)²), with τ = m/σ²ħ = 2mΔx(0)²/ħ.
Common mix-ups
- "σ is the width of the hump." σ is the spread in k. The hump's width is Δx(0) = 1/(σ√2), so a larger σ gives a narrower hump.
- "The packet spreads because probability leaks away." The area under P stays 1 (Norm reads 1.00 until the packet outgrows the ±40 window). The peak falls because the same area spreads wider.
- "Momentum holds a moving packet together." k₀ only moves the centre. The width law has no k₀ in it.
Things to try on your own
- Set k₀ = 1, σ = 1, m = 1. Predict the Time at which xpeak reaches 3, then check the readout.
- Set σ = 2, m = 5. Compute τ, then find the Time at which σ²|α|² has doubled from 0.250 to 0.500.
- Set k₀ = 0, σ = 1, m = 1. Using Δx(t) = Δx(0)√(1 + (t/τ)²), predict when Peak P has halved from 0.5642, then check.